PROJECT SUMMARY The long-term objective of this proposal is to understand the mechanisms responsible for loss of consciousness (LOC) under general anesthesia. Previous invasive studies in animal models have identified candidate mechanisms, but translating those findings to human consciousness remains approximate. More precise determinations of states of consciousness are feasible in human subjects, but non-invasive measures of brain activity (fMRI, EEG, MEG) can only indirectly assess the underlying neural circuitry. This proposal will overcome these limitations by taking advantage of the unique opportunity to directly record from the human brain. Using electrocorticography in neurosurgical patients, we will investigate neural networks modulated by general anesthesia. Our overarching goal is to identify patterns of activity and connectivity in cortical networks that track changes in contents of consciousness (i.e. awareness) under anesthesia and during sleep. Our approach is to identify changes in the networks underlying auditory predictive coding that occur upon LOC. Predictive coding minimizes the differences between internally generated constructs and empirical data, subserved by ongoing interaction between sensory and higher-order cortical regions. This model is ideal for this project because it engages the crucial interplay between predictions of the world and sensory observations of the world, a fundamental function of consciousness. The scientific premise of this project is that at a systems level, disruption of predictive coding subserved by large-scale cortical networks represents a signature of anesthetic-induced unconsciousness. To accomplish our goals, we will pursue three specific aims. The first aim seeks to refine our understanding of the cortical networks involved in auditory predictive coding in awake behaving subjects. This aim will focus on identifying the connectivity of the networks subserving predictive coding over short and long time scales, as effects of LOC on these networks are believed to be distinct. The second aim examines changes in network structure upon anesthesia LOC. This will be achieved by recording brain activity from subjects during induction of general anesthesia. The generality of the findings will be tested using two different anesthetic agents. Aim 3 seeks to identify common electrophysiological signatures for LOC under anesthesia and during sleep. This will be achieved by measuring brain activity in the same subjects during natural sleep. The results will have broad clinical applicability to defining and interpreting prognostic signs in patients with altered mental status (e.g. chronic vegetative and minimally conscious states), mental illness (e.g. delirium and schizophrenia) and development of novel algorithms for use in monitoring depth of anesthesia.